create-gety-connector
Pass
Audited by Gen Agent Trust Hub on Sep 29, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill facilitates the creation of connectors designed to ingest and index data from potentially untrusted external sources (APIs, SaaS applications, databases).
- Ingestion points: Data enters the agent's context through the
poll()lifecycle method implemented insrc/index.ts. - Boundary markers: The instructions recommend distinct handling for
markdownandplaintextcontent formats but do not explicitly mandate specific boundary markers or delimiters to isolate untrusted content from the rest of the document. - Capability inventory: Generated connectors utilize the
@gety-ai/connector-sdkto performupsertanddeloperations, modifying the host application's search index. - Sanitization: While the skill advises clarifying source semantics and document mapping, it does not provide explicit instructions for sanitizing or escaping the external content before it is indexed.
- [DYNAMIC_EXECUTION]: The development workflow involves generating code, compiling artifacts using
deno task build, and executing them via a local runner and the Gety application environment. This dynamic loading and execution of generated code is the intended primary purpose of the skill. - [EXTERNAL_DOWNLOADS]: The skill requires cloning the official template repository from GitHub (
gety-ai/gety-sample-connector) and references the vendor's own SDK (@gety-ai/connector-sdk). These resources originate from the vendor's official infrastructure.
Audit Metadata